Magnetic manipulation: Complex magnetic field processing leads to a new class of composite materials

On a computer screen in Jim Martin’s (1122) lab is a model of particles suspended in a solid polymer. They’re distributed randomly, to represent what industry achieves when it dumps particles into its product mixes to greatly improve the strength and durability of tires and of airplane and auto parts built out of toughened plastics.

How would the properties of particle composites benefit by having the particles organized into complex structures? And how expensive would it be to achieve this? Before Jim’s work, the answers of the materials community were, not much, and very.

Jim, a physical chemist turned computer modeler, circuit designer, and device developer, hits a key to initiate a simulation of the effect on the random particle suspension of his low-powered, newly developed device — three circular magnets, of approximate diameters 9, 6, and 3 inches, nested like Matrioshka dolls, except that each is aligned along either the breadth, depth or height (for the technically minded, the x, y or z axis) of the suspension placed at their center. The overall structure of these magnets — cutting directly across each others’ major axes — gives the feeling of a cubist sculpture, or of a junior version of the huge machine in the Carl Sagan-based movie Contact that takes the young woman scientist through an Einsteinian worm hole. But this is reality, not fiction, and this magnet is only about a cubic foot and it doesn’t spin, make noise, or send anyone into space (though Jim assures me he’s been working hard on that).

The arrangement creates some very remarkable effects that have led to an entirely new class of composites. Simply put, the magnetic fields lead to highly organized particle structures in the solidifying polymer resin. Electrically nonconductive suspensions are rendered conductive, magnetic properties are greatly enhanced, and solid composites show an unprecedented tendency to contract in magnetic fields.

First came magnetic watermark

Jim explains: "A few years back Bob Anderson (1843) and I started to investigate the benefits of using magnetic or electric fields to create polymer composites with embedded particles formed into chains — the kind of structures that iron filings form between the ends of a horseshoe magnet. We found we could increase the dielectric properties of ceramic particle composites severalfold, and with the help of Gene Venturini [1122] we found similar increases in the magnetic properties, most notably an increased tendency to magnetize in a magnetic field, and to retain a lot more of that magnetization when the field is removed. Retained magnetism is called remanence. This observation led to a patent on a magnetic watermark, with Anderson and Chris Tigges [1742], intended to discourage credit card theft."

Jim hits a button on his computer keyboard that starts a simulation of particles forming chains. The chains form quickly and then clump together to form columns.

The magnetic watermark should up the ante for identity thieves who use handheld scanners to record names, account numbers, and other identifying information from the magnetic stripes on a credit card. The information ordinarily could be downloaded onto a personal computer and used to make phony cards. But if the magnetic stripe on the credit card were embossed with a difficult-to-duplicate magnetic watermark, created by magnetic field structuring, the scam would be much harder to pull off. This nonerasable watermark would encode data only in variations of the magnetic remanence magnitude of the stripe, and is a layer of information in addition to the normal magnetization direction data.

Chains, strains, and a remarkable finding

"By structuring particles into chains we can also make composites conduct electricity, the current traveling through the chains," says Jim. "Remarkably, we find that this conduction is highly sensitive to any kind of swelling or deformation of the polymer. For example, a reversible 6 percent compressive strain can change the conductivity by 11 orders of magnitude, and a slight amount of swelling by a chemical vapor can change the conductivity by 10 orders of magnitude. So Bob Hughes [retired], Bob Anderson, and I obtained a Sandia patent on these materials as sensors.

"We soon tired of making particle chain composites and wanted to do something new. Working with T. Halsey, of the James Franck Institute at the University of Chicago, we discovered that it is possible to organize particles into layered sheetlike structures by placing suspensions in a field rotating in a plane, a so-called biaxial field. So now we could make chains and sheets, but that seemed to be all we could possibly make, since if we applied a field rotating in three dimensions, the magnetic interactions causing the particles to attract should vanish."

Jim shows me a simulation of sheet formation in a rotating magnetic field, explaining further, "The biaxial magnetic field interactions that give particle sheets are just the opposite of the interactions in a uniaxial field. So the sum of both — a triaxial field — would cancel all interactions. Kind of a fancy way to do nothing. But as it turned out we built a triaxial magnet anyway, as a way to use feedback to control the conduction of particle-based chemical sensors, an LDRD program developed with Rod Williamson [1835]. Building the triaxial magnet was not too hard, since we had already designed the rotating field magnet and so just had to add another coil."

"We were completely surprised when we discovered all kinds of interesting particle structures at the balance point of the triaxial magnet, where all three field amplitudes are equal. The lack of a ‘null’ effect was a complete surprise to us, and for a while we thought something was wrong with our magnet. But when we became convinced the fields were indeed balanced, long-time collaborator Bob Anderson and I got busy trying to understand what was going on."

Surprising, strange particle interactions

"It turns out that strange second-order particle interactions of surprising magnitude remained, interactions that are purely many-body effects, so that to determine how any two particles interact you have to know where all the other particles are. Really just as complicated and weird as things can be."

To illustrate the strange phenomena that can emerge, Jim takes me to his lab and turns on the triaxial magnet, which causes a particle sediment to blossom like a rose. Adjusting one field component frequency causes the rose to transform into a pulsing structure, something like the view inside a washing machine. Another adjustment causes a kind of controlled chaos — particle sheets that change orientation abruptly and discontinuously. One more tweak of a knob and the suspension explodes into whirling vortices that seem to be ready to head off for a trailer park back East.

All of this is interesting, but I can’t help wondering what can be done with this. "That’s the really neat part. When we apply the right kind of triaxial field all of the particles arrange themselves to optimize the magnetic properties of the suspension. But this also optimizes a lot of other properties, such as the thermal and electrical conductivity. So now we can make highly optimized particle composites. For example, a chemical sensor made with this technique has 1,000 times the conductivity of one made with a uniaxial field."

Jim shows me one composite made in his lab where the particles have been formed into a honeycomb structure. "Gerald Gulley [visiting faculty from Dominican University], Gene Venturini, and I spent several months studying the magnetic properties of these materials, and this honeycomb structure has the best magnetic properties of any sample we were able to make." I take his word on it and smile obligingly, but don’t know what to make of it.

A superfast artificial muscle?

Jim leads me over to an optical table and shows me the product of a current LDRD, a sort of rubber worm suspended in a magnet by the arm of a balance. "With this apparatus we are trying to develop a superfast artificial muscle, something Bob Anderson and I had theorized about a couple years ago. The idea is to fill a polymer with magnetic particles, stretch it, and see how much it contracts when you turn on the magnetic field. By using fields to optimize the magnetic properties of the composite we get a fivefold increase in the contraction, and we really believe we can still do better, much better than existing materials. We even believe we can develop artificial muscles that will remain contracted after a brief magnetic pulse, then release after another pulse."

The same materials could be used to make extremely sensitive sensor pads to give robots an imitation of the sensation afforded by the human sense of touch, Jim says. "Used in little soft finger pads, the harder the robot pushes, the more electricity is conducted."

At the heart of this research is the triaxial magnet. While it cost Jim about $20,000 to build it, the same machine custom built by professionals would cost at least $150,000, and would not be as good, says Boris Khusid, mechanical engineering professor at the New Jersey Institute of Technology in Newark.

Khusid says that although it is easy to make magnetic fields that oscillate at low frequencies, it is difficult to make magnetic fields at the roughly kilohertz frequencies needed, because the power demand is so high. "It could easily require 20,000 watts of power to drive just one coil," he says, "so Jim has connected each magnet coil to a capacitor bank to create a resonance where the energy just sloshes back and forth between the coil and the capacitor bank. Then, only an ordinary audio power supply is needed to compensate for the modest heat output of the coils. The problem with duplicating this system is that Jim has not published the design of the capacitor banks, which is the key to making this system work."

"Why should anyone pay $150,000 only to be able to try to duplicate Jim’s results poorly?" Khusid adds.

The design of the capacitor banks presented Jim with a major problem. "To be able to make each coil resonate over a broad frequency range required capacitor banks with essentially continuous tunability. To do this with a conventional design would have required a tremendous number of expensive capacitors for each coil. On a zoo outing with my kids it suddenly struck me how to build a much more efficient bank. This design needed only a dozen capacitors to give 354,000 capacitance values spanning three decades of capacitance. The bank has a fractal distribution of states, which is really what we needed. Jess Wilcoxon [1122] helped interface the bank to a computer, and Larry Shapnek [15322] did a beautiful job building the banks and ensuring we would not fry ourselves with the high voltages that occur."

A novel approach for novel materials

Says Khusid, "I strongly believe Jim’s methods will be widely used for producing composites as soon as his setup for generating multi-axial magnetic fields becomes commercially available. Jim’s studies originated a novel approach toward producing materials with tailored microstructures. His composites demonstrate unique properties and have the potential of enhancing sensors for a wide range of applications."

Back at the computer screen, a simulation shows the washing machine motion I had seen in the lab. The pulsing speeds up, and almost miraculously the particles form into a fibrillating honeycomb, as odd a birth as I have witnessed.

The structures on the computer screen then form in shapes no one has observed before, transforming in apparently unpredictable ways.

"In a triaxial magnetic field," says Jim, "things happen that are reminiscent of quantum physics, not classical physics, because of the strong many-body effects. For example, if one applies an attractive interaction to a popcorn ball of particles, these clumped particles should hold together. But in a triaxial magnetic field the clump explodes, six ‘arms’ shooting out like little mushroom clouds. Likewise, particle clusterings similar to organic molecules (ballecules?) are stable. It is a kind of magnetic chemistry with rules we still don’t fully understand." He changes the relative frequencies of the magnets; weird new structures result.

In fact, Jim says, "We can create a variety of particle structures that cannot be produced by any other known means, which has led to a patent application for this process."

Lucky and surprised

The potential byproducts, all unexpected, amuse Jim. "If we had been asked to make such complex composites we would have said it is impossible," he says. "So I think we are just lucky, and as surprised as anyone."

Says Paul Fleury, Dean of Engineering at Yale and a former Sandia research VP, "I think the work is scientifically very creative and interesting. Jim has devised a controllable model system that can be used to explore a number of structural configurations as well as magnetic states. . . . As to commercial application, it is quite early æ but there may be possibilities of scaling up so as to control magnetic and other microstructures with the large enhancements in susceptibility that he has achieved in the lab."

Khusid adds that Jim also uses the multi-axial fields "as a [science] tool to study the elusive nature of many-body effects in systems governed by the long-range particle interactions. . . ."

The results are discussed in detail in papers for the Journal of Chemical Physics (Jan. 15, 2003), written by Jim with Bob Anderson and Rod Williamson, and in Physical Review B/Condensed Matter and Materials Physics (March 2003) with Bob, Judy Odinek (2522), Douglas Adolf (1811), and Jennifer Williamson (past summer student).

The work was supported by DOE’s Division of Materials Sciences and Engineering, Office of Basic Energy Sciences.